2 * software RGB to RGB converter
3 * pluralize by software PAL8 to RGB converter
4 * software YUV to YUV converter
5 * software YUV to RGB converter
6 * Written by Nick Kurshev.
7 * palette & YUV & runtime CPU stuff by Michael (michaelni@gmx.at)
8 * lot of big-endian byte order fixes by Alex Beregszaszi
10 * This file is part of FFmpeg.
12 * FFmpeg is free software; you can redistribute it and/or
13 * modify it under the terms of the GNU Lesser General Public
14 * License as published by the Free Software Foundation; either
15 * version 2.1 of the License, or (at your option) any later version.
17 * FFmpeg is distributed in the hope that it will be useful,
18 * but WITHOUT ANY WARRANTY; without even the implied warranty of
19 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
20 * Lesser General Public License for more details.
22 * You should have received a copy of the GNU Lesser General Public
23 * License along with FFmpeg; if not, write to the Free Software
24 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
29 #include "libavutil/attributes.h"
31 static inline void rgb24tobgr32_c(const uint8_t *src
, uint8_t *dst
,
35 const uint8_t *s
= src
;
36 const uint8_t *end
= s
+ src_size
;
40 /* RGB24 (= R, G, B) -> RGB32 (= A, B, G, R) */
55 static inline void rgb32tobgr24_c(const uint8_t *src
, uint8_t *dst
,
59 const uint8_t *s
= src
;
60 const uint8_t *end
= s
+ src_size
;
64 /* RGB32 (= A, B, G, R) -> RGB24 (= R, G, B) */
80 * original by Strepto/Astral
81 * ported to gcc & bugfixed: A'rpi
82 * MMXEXT, 3DNOW optimization by Nick Kurshev
83 * 32-bit C version, and and&add trick by Michael Niedermayer
85 static inline void rgb15to16_c(const uint8_t *src
, uint8_t *dst
, int src_size
)
87 register uint8_t *d
= dst
;
88 register const uint8_t *s
= src
;
89 register const uint8_t *end
= s
+ src_size
;
90 const uint8_t *mm_end
= end
- 3;
93 register unsigned x
= *((const uint32_t *)s
);
94 *((uint32_t *)d
) = (x
& 0x7FFF7FFF) + (x
& 0x7FE07FE0);
99 register unsigned short x
= *((const uint16_t *)s
);
100 *((uint16_t *)d
) = (x
& 0x7FFF) + (x
& 0x7FE0);
104 static inline void rgb16to15_c(const uint8_t *src
, uint8_t *dst
, int src_size
)
106 register uint8_t *d
= dst
;
107 register const uint8_t *s
= src
;
108 register const uint8_t *end
= s
+ src_size
;
109 const uint8_t *mm_end
= end
- 3;
112 register uint32_t x
= *((const uint32_t *)s
);
113 *((uint32_t *)d
) = ((x
>> 1) & 0x7FE07FE0) | (x
& 0x001F001F);
118 register uint16_t x
= *((const uint16_t *)s
);
119 *((uint16_t *)d
) = ((x
>> 1) & 0x7FE0) | (x
& 0x001F);
123 static inline void rgb32to16_c(const uint8_t *src
, uint8_t *dst
, int src_size
)
125 uint16_t *d
= (uint16_t *)dst
;
126 const uint8_t *s
= src
;
127 const uint8_t *end
= s
+ src_size
;
130 register int rgb
= *(const uint32_t *)s
;
132 *d
++ = ((rgb
& 0xFF) >> 3) +
133 ((rgb
& 0xFC00) >> 5) +
134 ((rgb
& 0xF80000) >> 8);
138 static inline void rgb32tobgr16_c(const uint8_t *src
, uint8_t *dst
,
141 uint16_t *d
= (uint16_t *)dst
;
142 const uint8_t *s
= src
;
143 const uint8_t *end
= s
+ src_size
;
146 register int rgb
= *(const uint32_t *)s
;
148 *d
++ = ((rgb
& 0xF8) << 8) +
149 ((rgb
& 0xFC00) >> 5) +
150 ((rgb
& 0xF80000) >> 19);
154 static inline void rgb32to15_c(const uint8_t *src
, uint8_t *dst
, int src_size
)
156 uint16_t *d
= (uint16_t *)dst
;
157 const uint8_t *s
= src
;
158 const uint8_t *end
= s
+ src_size
;
161 register int rgb
= *(const uint32_t *)s
;
163 *d
++ = ((rgb
& 0xFF) >> 3) +
164 ((rgb
& 0xF800) >> 6) +
165 ((rgb
& 0xF80000) >> 9);
169 static inline void rgb32tobgr15_c(const uint8_t *src
, uint8_t *dst
,
172 uint16_t *d
= (uint16_t *)dst
;
173 const uint8_t *s
= src
;
174 const uint8_t *end
= s
+ src_size
;
177 register int rgb
= *(const uint32_t *)s
;
179 *d
++ = ((rgb
& 0xF8) << 7) +
180 ((rgb
& 0xF800) >> 6) +
181 ((rgb
& 0xF80000) >> 19);
185 static inline void rgb24tobgr16_c(const uint8_t *src
, uint8_t *dst
,
188 uint16_t *d
= (uint16_t *)dst
;
189 const uint8_t *s
= src
;
190 const uint8_t *end
= s
+ src_size
;
196 *d
++ = (b
>> 3) | ((g
& 0xFC) << 3) | ((r
& 0xF8) << 8);
200 static inline void rgb24to16_c(const uint8_t *src
, uint8_t *dst
, int src_size
)
202 uint16_t *d
= (uint16_t *)dst
;
203 const uint8_t *s
= src
;
204 const uint8_t *end
= s
+ src_size
;
210 *d
++ = (b
>> 3) | ((g
& 0xFC) << 3) | ((r
& 0xF8) << 8);
214 static inline void rgb24tobgr15_c(const uint8_t *src
, uint8_t *dst
,
217 uint16_t *d
= (uint16_t *)dst
;
218 const uint8_t *s
= src
;
219 const uint8_t *end
= s
+ src_size
;
225 *d
++ = (b
>> 3) | ((g
& 0xF8) << 2) | ((r
& 0xF8) << 7);
229 static inline void rgb24to15_c(const uint8_t *src
, uint8_t *dst
, int src_size
)
231 uint16_t *d
= (uint16_t *)dst
;
232 const uint8_t *s
= src
;
233 const uint8_t *end
= s
+ src_size
;
239 *d
++ = (b
>> 3) | ((g
& 0xF8) << 2) | ((r
& 0xF8) << 7);
243 static inline void rgb15tobgr24_c(const uint8_t *src
, uint8_t *dst
,
247 const uint16_t *s
= (const uint16_t *)src
;
248 const uint16_t *end
= s
+ src_size
/ 2;
251 register uint16_t bgr
= *s
++;
252 *d
++ = ((bgr
&0x001F)<<3) | ((bgr
&0x001F)>> 2);
253 *d
++ = ((bgr
&0x03E0)>>2) | ((bgr
&0x03E0)>> 7);
254 *d
++ = ((bgr
&0x7C00)>>7) | ((bgr
&0x7C00)>>12);
258 static inline void rgb16tobgr24_c(const uint8_t *src
, uint8_t *dst
,
261 uint8_t *d
= (uint8_t *)dst
;
262 const uint16_t *s
= (const uint16_t *)src
;
263 const uint16_t *end
= s
+ src_size
/ 2;
266 register uint16_t bgr
= *s
++;
267 *d
++ = ((bgr
&0x001F)<<3) | ((bgr
&0x001F)>> 2);
268 *d
++ = ((bgr
&0x07E0)>>3) | ((bgr
&0x07E0)>> 9);
269 *d
++ = ((bgr
&0xF800)>>8) | ((bgr
&0xF800)>>13);
273 static inline void rgb15to32_c(const uint8_t *src
, uint8_t *dst
, int src_size
)
276 const uint16_t *s
= (const uint16_t *)src
;
277 const uint16_t *end
= s
+ src_size
/ 2;
280 register uint16_t bgr
= *s
++;
283 *d
++ = ((bgr
&0x7C00)>>7) | ((bgr
&0x7C00)>>12);
284 *d
++ = ((bgr
&0x03E0)>>2) | ((bgr
&0x03E0)>> 7);
285 *d
++ = ((bgr
&0x001F)<<3) | ((bgr
&0x001F)>> 2);
287 *d
++ = ((bgr
&0x001F)<<3) | ((bgr
&0x001F)>> 2);
288 *d
++ = ((bgr
&0x03E0)>>2) | ((bgr
&0x03E0)>> 7);
289 *d
++ = ((bgr
&0x7C00)>>7) | ((bgr
&0x7C00)>>12);
295 static inline void rgb16to32_c(const uint8_t *src
, uint8_t *dst
, int src_size
)
298 const uint16_t *s
= (const uint16_t *)src
;
299 const uint16_t *end
= s
+ src_size
/ 2;
302 register uint16_t bgr
= *s
++;
305 *d
++ = ((bgr
&0xF800)>>8) | ((bgr
&0xF800)>>13);
306 *d
++ = ((bgr
&0x07E0)>>3) | ((bgr
&0x07E0)>> 9);
307 *d
++ = ((bgr
&0x001F)<<3) | ((bgr
&0x001F)>> 2);
309 *d
++ = ((bgr
&0x001F)<<3) | ((bgr
&0x001F)>> 2);
310 *d
++ = ((bgr
&0x07E0)>>3) | ((bgr
&0x07E0)>> 9);
311 *d
++ = ((bgr
&0xF800)>>8) | ((bgr
&0xF800)>>13);
317 static inline void shuffle_bytes_2103_c(const uint8_t *src
, uint8_t *dst
,
320 int idx
= 15 - src_size
;
321 const uint8_t *s
= src
- idx
;
322 uint8_t *d
= dst
- idx
;
324 for (; idx
< 15; idx
+= 4) {
325 register int v
= *(const uint32_t *)&s
[idx
], g
= v
& 0xff00ff00;
327 *(uint32_t *)&d
[idx
] = (v
>> 16) + g
+ (v
<< 16);
331 static inline void rgb24tobgr24_c(const uint8_t *src
, uint8_t *dst
, int src_size
)
335 for (i
= 0; i
< src_size
; i
+= 3) {
336 register uint8_t x
= src
[i
+ 2];
337 dst
[i
+ 1] = src
[i
+ 1];
338 dst
[i
+ 2] = src
[i
+ 0];
343 static inline void yuvPlanartoyuy2_c(const uint8_t *ysrc
, const uint8_t *usrc
,
344 const uint8_t *vsrc
, uint8_t *dst
,
345 int width
, int height
,
346 int lumStride
, int chromStride
,
347 int dstStride
, int vertLumPerChroma
)
350 const int chromWidth
= width
>> 1;
352 for (y
= 0; y
< height
; y
++) {
354 uint64_t *ldst
= (uint64_t *)dst
;
355 const uint8_t *yc
= ysrc
, *uc
= usrc
, *vc
= vsrc
;
356 for (i
= 0; i
< chromWidth
; i
+= 2) {
357 uint64_t k
= yc
[0] + (uc
[0] << 8) +
358 (yc
[1] << 16) + ((unsigned) vc
[0] << 24);
359 uint64_t l
= yc
[2] + (uc
[1] << 8) +
360 (yc
[3] << 16) + ((unsigned) vc
[1] << 24);
361 *ldst
++ = k
+ (l
<< 32);
368 int *idst
= (int32_t *)dst
;
369 const uint8_t *yc
= ysrc
, *uc
= usrc
, *vc
= vsrc
;
371 for (i
= 0; i
< chromWidth
; i
++) {
373 *idst
++ = (yc
[0] << 24) + (uc
[0] << 16) +
374 (yc
[1] << 8) + (vc
[0] << 0);
376 *idst
++ = yc
[0] + (uc
[0] << 8) +
377 (yc
[1] << 16) + (vc
[0] << 24);
384 if ((y
& (vertLumPerChroma
- 1)) == vertLumPerChroma
- 1) {
394 * Height should be a multiple of 2 and width should be a multiple of 16.
395 * (If this is a problem for anyone then tell me, and I will fix it.)
397 static inline void yv12toyuy2_c(const uint8_t *ysrc
, const uint8_t *usrc
,
398 const uint8_t *vsrc
, uint8_t *dst
,
399 int width
, int height
, int lumStride
,
400 int chromStride
, int dstStride
)
402 //FIXME interpolate chroma
403 yuvPlanartoyuy2_c(ysrc
, usrc
, vsrc
, dst
, width
, height
, lumStride
,
404 chromStride
, dstStride
, 2);
407 static inline void yuvPlanartouyvy_c(const uint8_t *ysrc
, const uint8_t *usrc
,
408 const uint8_t *vsrc
, uint8_t *dst
,
409 int width
, int height
,
410 int lumStride
, int chromStride
,
411 int dstStride
, int vertLumPerChroma
)
414 const int chromWidth
= width
>> 1;
416 for (y
= 0; y
< height
; y
++) {
418 uint64_t *ldst
= (uint64_t *)dst
;
419 const uint8_t *yc
= ysrc
, *uc
= usrc
, *vc
= vsrc
;
420 for (i
= 0; i
< chromWidth
; i
+= 2) {
421 uint64_t k
= uc
[0] + (yc
[0] << 8) +
422 (vc
[0] << 16) + ((unsigned) yc
[1] << 24);
423 uint64_t l
= uc
[1] + (yc
[2] << 8) +
424 (vc
[1] << 16) + ((unsigned) yc
[3] << 24);
425 *ldst
++ = k
+ (l
<< 32);
432 int *idst
= (int32_t *)dst
;
433 const uint8_t *yc
= ysrc
, *uc
= usrc
, *vc
= vsrc
;
435 for (i
= 0; i
< chromWidth
; i
++) {
437 *idst
++ = (uc
[0] << 24) + (yc
[0] << 16) +
438 (vc
[0] << 8) + (yc
[1] << 0);
440 *idst
++ = uc
[0] + (yc
[0] << 8) +
441 (vc
[0] << 16) + (yc
[1] << 24);
448 if ((y
& (vertLumPerChroma
- 1)) == vertLumPerChroma
- 1) {
458 * Height should be a multiple of 2 and width should be a multiple of 16
459 * (If this is a problem for anyone then tell me, and I will fix it.)
461 static inline void yv12touyvy_c(const uint8_t *ysrc
, const uint8_t *usrc
,
462 const uint8_t *vsrc
, uint8_t *dst
,
463 int width
, int height
, int lumStride
,
464 int chromStride
, int dstStride
)
466 //FIXME interpolate chroma
467 yuvPlanartouyvy_c(ysrc
, usrc
, vsrc
, dst
, width
, height
, lumStride
,
468 chromStride
, dstStride
, 2);
472 * Width should be a multiple of 16.
474 static inline void yuv422ptouyvy_c(const uint8_t *ysrc
, const uint8_t *usrc
,
475 const uint8_t *vsrc
, uint8_t *dst
,
476 int width
, int height
, int lumStride
,
477 int chromStride
, int dstStride
)
479 yuvPlanartouyvy_c(ysrc
, usrc
, vsrc
, dst
, width
, height
, lumStride
,
480 chromStride
, dstStride
, 1);
484 * Width should be a multiple of 16.
486 static inline void yuv422ptoyuy2_c(const uint8_t *ysrc
, const uint8_t *usrc
,
487 const uint8_t *vsrc
, uint8_t *dst
,
488 int width
, int height
, int lumStride
,
489 int chromStride
, int dstStride
)
491 yuvPlanartoyuy2_c(ysrc
, usrc
, vsrc
, dst
, width
, height
, lumStride
,
492 chromStride
, dstStride
, 1);
496 * Height should be a multiple of 2 and width should be a multiple of 16.
497 * (If this is a problem for anyone then tell me, and I will fix it.)
499 static inline void yuy2toyv12_c(const uint8_t *src
, uint8_t *ydst
,
500 uint8_t *udst
, uint8_t *vdst
,
501 int width
, int height
, int lumStride
,
502 int chromStride
, int srcStride
)
505 const int chromWidth
= width
>> 1;
507 for (y
= 0; y
< height
; y
+= 2) {
509 for (i
= 0; i
< chromWidth
; i
++) {
510 ydst
[2 * i
+ 0] = src
[4 * i
+ 0];
511 udst
[i
] = src
[4 * i
+ 1];
512 ydst
[2 * i
+ 1] = src
[4 * i
+ 2];
513 vdst
[i
] = src
[4 * i
+ 3];
518 for (i
= 0; i
< chromWidth
; i
++) {
519 ydst
[2 * i
+ 0] = src
[4 * i
+ 0];
520 ydst
[2 * i
+ 1] = src
[4 * i
+ 2];
529 static inline void planar2x_c(const uint8_t *src
, uint8_t *dst
, int srcWidth
,
530 int srcHeight
, int srcStride
, int dstStride
)
537 for (x
= 0; x
< srcWidth
- 1; x
++) {
538 dst
[2 * x
+ 1] = (3 * src
[x
] + src
[x
+ 1]) >> 2;
539 dst
[2 * x
+ 2] = (src
[x
] + 3 * src
[x
+ 1]) >> 2;
541 dst
[2 * srcWidth
- 1] = src
[srcWidth
- 1];
545 for (y
= 1; y
< srcHeight
; y
++) {
546 const int mmxSize
= 1;
548 dst
[0] = (src
[0] * 3 + src
[srcStride
]) >> 2;
549 dst
[dstStride
] = (src
[0] + 3 * src
[srcStride
]) >> 2;
551 for (x
= mmxSize
- 1; x
< srcWidth
- 1; x
++) {
552 dst
[2 * x
+ 1] = (src
[x
+ 0] * 3 + src
[x
+ srcStride
+ 1]) >> 2;
553 dst
[2 * x
+ dstStride
+ 2] = (src
[x
+ 0] + 3 * src
[x
+ srcStride
+ 1]) >> 2;
554 dst
[2 * x
+ dstStride
+ 1] = (src
[x
+ 1] + 3 * src
[x
+ srcStride
]) >> 2;
555 dst
[2 * x
+ 2] = (src
[x
+ 1] * 3 + src
[x
+ srcStride
]) >> 2;
557 dst
[srcWidth
* 2 - 1] = (src
[srcWidth
- 1] * 3 + src
[srcWidth
- 1 + srcStride
]) >> 2;
558 dst
[srcWidth
* 2 - 1 + dstStride
] = (src
[srcWidth
- 1] + 3 * src
[srcWidth
- 1 + srcStride
]) >> 2;
560 dst
+= dstStride
* 2;
567 for (x
= 0; x
< srcWidth
- 1; x
++) {
568 dst
[2 * x
+ 1] = (src
[x
] * 3 + src
[x
+ 1]) >> 2;
569 dst
[2 * x
+ 2] = (src
[x
] + 3 * src
[x
+ 1]) >> 2;
571 dst
[2 * srcWidth
- 1] = src
[srcWidth
- 1];
575 * Height should be a multiple of 2 and width should be a multiple of 16.
576 * (If this is a problem for anyone then tell me, and I will fix it.)
577 * Chrominance data is only taken from every second line, others are ignored.
578 * FIXME: Write HQ version.
580 static inline void uyvytoyv12_c(const uint8_t *src
, uint8_t *ydst
,
581 uint8_t *udst
, uint8_t *vdst
,
582 int width
, int height
, int lumStride
,
583 int chromStride
, int srcStride
)
586 const int chromWidth
= width
>> 1;
588 for (y
= 0; y
< height
; y
+= 2) {
590 for (i
= 0; i
< chromWidth
; i
++) {
591 udst
[i
] = src
[4 * i
+ 0];
592 ydst
[2 * i
+ 0] = src
[4 * i
+ 1];
593 vdst
[i
] = src
[4 * i
+ 2];
594 ydst
[2 * i
+ 1] = src
[4 * i
+ 3];
599 for (i
= 0; i
< chromWidth
; i
++) {
600 ydst
[2 * i
+ 0] = src
[4 * i
+ 1];
601 ydst
[2 * i
+ 1] = src
[4 * i
+ 3];
611 * Height should be a multiple of 2 and width should be a multiple of 2.
612 * (If this is a problem for anyone then tell me, and I will fix it.)
613 * Chrominance data is only taken from every second line,
614 * others are ignored in the C version.
615 * FIXME: Write HQ version.
617 void ff_rgb24toyv12_c(const uint8_t *src
, uint8_t *ydst
, uint8_t *udst
,
618 uint8_t *vdst
, int width
, int height
, int lumStride
,
619 int chromStride
, int srcStride
, int32_t *rgb2yuv
)
621 int32_t ry
= rgb2yuv
[RY_IDX
], gy
= rgb2yuv
[GY_IDX
], by
= rgb2yuv
[BY_IDX
];
622 int32_t ru
= rgb2yuv
[RU_IDX
], gu
= rgb2yuv
[GU_IDX
], bu
= rgb2yuv
[BU_IDX
];
623 int32_t rv
= rgb2yuv
[RV_IDX
], gv
= rgb2yuv
[GV_IDX
], bv
= rgb2yuv
[BV_IDX
];
625 const int chromWidth
= width
>> 1;
627 for (y
= 0; y
< height
; y
+= 2) {
629 for (i
= 0; i
< chromWidth
; i
++) {
630 unsigned int b
= src
[6 * i
+ 0];
631 unsigned int g
= src
[6 * i
+ 1];
632 unsigned int r
= src
[6 * i
+ 2];
634 unsigned int Y
= ((ry
* r
+ gy
* g
+ by
* b
) >> RGB2YUV_SHIFT
) + 16;
635 unsigned int V
= ((rv
* r
+ gv
* g
+ bv
* b
) >> RGB2YUV_SHIFT
) + 128;
636 unsigned int U
= ((ru
* r
+ gu
* g
+ bu
* b
) >> RGB2YUV_SHIFT
) + 128;
646 Y
= ((ry
* r
+ gy
* g
+ by
* b
) >> RGB2YUV_SHIFT
) + 16;
655 for (i
= 0; i
< chromWidth
; i
++) {
656 unsigned int b
= src
[6 * i
+ 0];
657 unsigned int g
= src
[6 * i
+ 1];
658 unsigned int r
= src
[6 * i
+ 2];
660 unsigned int Y
= ((ry
* r
+ gy
* g
+ by
* b
) >> RGB2YUV_SHIFT
) + 16;
668 Y
= ((ry
* r
+ gy
* g
+ by
* b
) >> RGB2YUV_SHIFT
) + 16;
678 static void interleaveBytes_c(const uint8_t *src1
, const uint8_t *src2
,
679 uint8_t *dest
, int width
, int height
,
680 int src1Stride
, int src2Stride
, int dstStride
)
684 for (h
= 0; h
< height
; h
++) {
686 for (w
= 0; w
< width
; w
++) {
687 dest
[2 * w
+ 0] = src1
[w
];
688 dest
[2 * w
+ 1] = src2
[w
];
696 static void deinterleaveBytes_c(const uint8_t *src
, uint8_t *dst1
, uint8_t *dst2
,
697 int width
, int height
, int srcStride
,
698 int dst1Stride
, int dst2Stride
)
702 for (h
= 0; h
< height
; h
++) {
704 for (w
= 0; w
< width
; w
++) {
705 dst1
[w
] = src
[2 * w
+ 0];
706 dst2
[w
] = src
[2 * w
+ 1];
714 static inline void vu9_to_vu12_c(const uint8_t *src1
, const uint8_t *src2
,
715 uint8_t *dst1
, uint8_t *dst2
,
716 int width
, int height
,
717 int srcStride1
, int srcStride2
,
718 int dstStride1
, int dstStride2
)
724 for (y
= 0; y
< h
; y
++) {
725 const uint8_t *s1
= src1
+ srcStride1
* (y
>> 1);
726 uint8_t *d
= dst1
+ dstStride1
* y
;
727 for (x
= 0; x
< w
; x
++)
728 d
[2 * x
] = d
[2 * x
+ 1] = s1
[x
];
730 for (y
= 0; y
< h
; y
++) {
731 const uint8_t *s2
= src2
+ srcStride2
* (y
>> 1);
732 uint8_t *d
= dst2
+ dstStride2
* y
;
733 for (x
= 0; x
< w
; x
++)
734 d
[2 * x
] = d
[2 * x
+ 1] = s2
[x
];
738 static inline void yvu9_to_yuy2_c(const uint8_t *src1
, const uint8_t *src2
,
739 const uint8_t *src3
, uint8_t *dst
,
740 int width
, int height
,
741 int srcStride1
, int srcStride2
,
742 int srcStride3
, int dstStride
)
748 for (y
= 0; y
< h
; y
++) {
749 const uint8_t *yp
= src1
+ srcStride1
* y
;
750 const uint8_t *up
= src2
+ srcStride2
* (y
>> 2);
751 const uint8_t *vp
= src3
+ srcStride3
* (y
>> 2);
752 uint8_t *d
= dst
+ dstStride
* y
;
753 for (x
= 0; x
< w
; x
++) {
754 const int x2
= x
<< 2;
755 d
[8 * x
+ 0] = yp
[x2
];
756 d
[8 * x
+ 1] = up
[x
];
757 d
[8 * x
+ 2] = yp
[x2
+ 1];
758 d
[8 * x
+ 3] = vp
[x
];
759 d
[8 * x
+ 4] = yp
[x2
+ 2];
760 d
[8 * x
+ 5] = up
[x
];
761 d
[8 * x
+ 6] = yp
[x2
+ 3];
762 d
[8 * x
+ 7] = vp
[x
];
767 static void extract_even_c(const uint8_t *src
, uint8_t *dst
, int count
)
773 dst
[count
] = src
[2 * count
];
778 static void extract_even2_c(const uint8_t *src
, uint8_t *dst0
, uint8_t *dst1
,
786 dst0
[count
] = src
[4 * count
+ 0];
787 dst1
[count
] = src
[4 * count
+ 2];
792 static void extract_even2avg_c(const uint8_t *src0
, const uint8_t *src1
,
793 uint8_t *dst0
, uint8_t *dst1
, int count
)
801 dst0
[count
] = (src0
[4 * count
+ 0] + src1
[4 * count
+ 0]) >> 1;
802 dst1
[count
] = (src0
[4 * count
+ 2] + src1
[4 * count
+ 2]) >> 1;
807 static void extract_odd2_c(const uint8_t *src
, uint8_t *dst0
, uint8_t *dst1
,
816 dst0
[count
] = src
[4 * count
+ 0];
817 dst1
[count
] = src
[4 * count
+ 2];
822 static void extract_odd2avg_c(const uint8_t *src0
, const uint8_t *src1
,
823 uint8_t *dst0
, uint8_t *dst1
, int count
)
833 dst0
[count
] = (src0
[4 * count
+ 0] + src1
[4 * count
+ 0]) >> 1;
834 dst1
[count
] = (src0
[4 * count
+ 2] + src1
[4 * count
+ 2]) >> 1;
839 static void yuyvtoyuv420_c(uint8_t *ydst
, uint8_t *udst
, uint8_t *vdst
,
840 const uint8_t *src
, int width
, int height
,
841 int lumStride
, int chromStride
, int srcStride
)
844 const int chromWidth
= FF_CEIL_RSHIFT(width
, 1);
846 for (y
= 0; y
< height
; y
++) {
847 extract_even_c(src
, ydst
, width
);
849 extract_odd2avg_c(src
- srcStride
, src
, udst
, vdst
, chromWidth
);
859 static void yuyvtoyuv422_c(uint8_t *ydst
, uint8_t *udst
, uint8_t *vdst
,
860 const uint8_t *src
, int width
, int height
,
861 int lumStride
, int chromStride
, int srcStride
)
864 const int chromWidth
= FF_CEIL_RSHIFT(width
, 1);
866 for (y
= 0; y
< height
; y
++) {
867 extract_even_c(src
, ydst
, width
);
868 extract_odd2_c(src
, udst
, vdst
, chromWidth
);
877 static void uyvytoyuv420_c(uint8_t *ydst
, uint8_t *udst
, uint8_t *vdst
,
878 const uint8_t *src
, int width
, int height
,
879 int lumStride
, int chromStride
, int srcStride
)
882 const int chromWidth
= FF_CEIL_RSHIFT(width
, 1);
884 for (y
= 0; y
< height
; y
++) {
885 extract_even_c(src
+ 1, ydst
, width
);
887 extract_even2avg_c(src
- srcStride
, src
, udst
, vdst
, chromWidth
);
897 static void uyvytoyuv422_c(uint8_t *ydst
, uint8_t *udst
, uint8_t *vdst
,
898 const uint8_t *src
, int width
, int height
,
899 int lumStride
, int chromStride
, int srcStride
)
902 const int chromWidth
= FF_CEIL_RSHIFT(width
, 1);
904 for (y
= 0; y
< height
; y
++) {
905 extract_even_c(src
+ 1, ydst
, width
);
906 extract_even2_c(src
, udst
, vdst
, chromWidth
);
915 static av_cold
void rgb2rgb_init_c(void)
917 rgb15to16
= rgb15to16_c
;
918 rgb15tobgr24
= rgb15tobgr24_c
;
919 rgb15to32
= rgb15to32_c
;
920 rgb16tobgr24
= rgb16tobgr24_c
;
921 rgb16to32
= rgb16to32_c
;
922 rgb16to15
= rgb16to15_c
;
923 rgb24tobgr16
= rgb24tobgr16_c
;
924 rgb24tobgr15
= rgb24tobgr15_c
;
925 rgb24tobgr32
= rgb24tobgr32_c
;
926 rgb32to16
= rgb32to16_c
;
927 rgb32to15
= rgb32to15_c
;
928 rgb32tobgr24
= rgb32tobgr24_c
;
929 rgb24to15
= rgb24to15_c
;
930 rgb24to16
= rgb24to16_c
;
931 rgb24tobgr24
= rgb24tobgr24_c
;
932 shuffle_bytes_2103
= shuffle_bytes_2103_c
;
933 rgb32tobgr16
= rgb32tobgr16_c
;
934 rgb32tobgr15
= rgb32tobgr15_c
;
935 yv12toyuy2
= yv12toyuy2_c
;
936 yv12touyvy
= yv12touyvy_c
;
937 yuv422ptoyuy2
= yuv422ptoyuy2_c
;
938 yuv422ptouyvy
= yuv422ptouyvy_c
;
939 yuy2toyv12
= yuy2toyv12_c
;
940 planar2x
= planar2x_c
;
941 ff_rgb24toyv12
= ff_rgb24toyv12_c
;
942 interleaveBytes
= interleaveBytes_c
;
943 deinterleaveBytes
= deinterleaveBytes_c
;
944 vu9_to_vu12
= vu9_to_vu12_c
;
945 yvu9_to_yuy2
= yvu9_to_yuy2_c
;
947 uyvytoyuv420
= uyvytoyuv420_c
;
948 uyvytoyuv422
= uyvytoyuv422_c
;
949 yuyvtoyuv420
= yuyvtoyuv420_c
;
950 yuyvtoyuv422
= yuyvtoyuv422_c
;